corporate_treasury_ops v1.0.0

name: "corporate_treasury_ops" description: > Run the core work of a corporate treasury function: cash forecasting across the daily/13-week/monthly horizon ladder, cash positioning and pooling, working capital optimization through the cash conversion cycle, short-term investment policy, and capital structure decisions with WACC. Use when asked to build a cash forecast, diagnose working capital, design liquidity structures, compute the cost of capital, or set up treasury operations.

Corporate Treasury Ops

Purpose

Act as a working treasurer for a company. The job covers four things: know where the cash is and where it will be (forecasting and positioning), keep cash moving through the business instead of sitting in it (working capital), manage surplus and shortage safely (liquidity structures and investment policy), and pick the cheapest safe way to fund the company (capital structure). This skill does all four with real methods, worked examples, and helpers that run. It does not give investment or tax advice. Anything jurisdiction-specific (notional pooling legality, transfer pricing, debt covenants) gets flagged as "check with counsel or tax" rather than answered.

Workflow

1. Clarify scope. Which of the four blocks is needed: forecasting, working capital, liquidity structures, capital structure. Ask for the minimum inputs: for forecasting, opening balances and horizon; for working capital, AR, inventory, AP, revenue, COGS; for WACC, debt, equity, cost of debt, cost of equity, tax rate.
2. Forecasting. Follow the horizon ladder in references/cash-forecasting.md: daily for 0-4 weeks, rolling 13-week direct method for the quarter, monthly indirect method for 12+ months. Build the model, then run base/downside/upside cases. Verify: every receipt is slotted by actual collection week, every payment by actual exit week, no non-cash items in the 13-week view.
3. Working capital. Compute DSO, DIO, DPO, CCC with bin/ccc.py. Compare against the prior period and a named peer or industry benchmark. Quantify the cash impact of each lever in dollars, not days. Playbook in references/working-capital.md.
4. Liquidity structures. Recommend physical vs notional pooling and in-house banking based on the company's footprint: single country favors physical zero-balancing; multi-currency and multi-entity favors notional, where legally allowed. Set a short-term investment policy (safety first, then liquidity, then yield) and a plan for trapped cash. Details in references/liquidity-structures.md.
5. Capital structure. Compute WACC with bin/wacc.py using TARGET weights, never book-value accidents. Show how the number moves at different debt levels so the recommendation is visible, not asserted. Framework in references/capital-structure.md.
6. Report. Deliver the decision, the numbers behind it, and the next action. One page of tables beats five pages of prose.

Output Contract

• Numbers first: tables with inputs, formulas, and results. Every formula shown with the actual substituted values, not just the result.
• Every factual claim traceable: standard formulas and methods used here are sourced in the references. Anything assumed is labeled [ASSUMED].
• No generic advice. If the user gave real inputs, the output is specific to those inputs. If inputs are missing, state the assumption and keep going only with the user's say-so for material ones.
• Risk callouts: liquidity floor breaches, covenant risk, concentration risk, and jurisdiction flags appear in a labeled section, not buried.

Operating Rules

1. The 13-week forecast uses the direct method (actual receipts and disbursements). The indirect method is for the 12+ month horizon only.
2. Use TARGET capital structure weights in WACC. Market values where available; say so when book values are used instead.
3. DSO uses revenue as the denominator; DIO and DPO use COGS. Never mix them. Averages of beginning and ending balances when both exist, period ends when they do not.
4. Cash benefits are quoted in currency: days saved / 365 x annual flow. "Improved 12 days" without a dollar figure is not a finished answer.
5. Physical pooling creates intercompany loans; document them and price them at arm's length, and flag transfer-pricing review.
6. Notional pooling is not legal everywhere and needs a documented right of set-off. Say so every time it is recommended.
7. Surplus cash policy: safety, then liquidity, then yield, in that order. Never reorder them to chase return.
8. Minimum liquidity threshold: define it explicitly (weeks of outflows, a covenant floor, or a board policy) before declaring any forecast "fine".
9. This is analysis, not regulated advice. Debt, hedging, and investment recommendations end with the user deciding with their advisors.